The recent lead-in-water crisis in Newark has renewed concerns about the crisis being a widespread problem in the nation. Using data on the exact home addresses of pregnant women residing in the city combined with information on the spatial boundary separating areas within the city serviced by two water treatment plants, we exploit an exogenous change in water chemistry that resulted in lead leaching into the tap water of one plant's service area, but not the other's, to identify a causal effect of prenatal lead exposure on fetal health. We find robust evidence of adverse health impacts, which has important policy implications in light of the substantial number of lead water pipes that remain in use as part of our aging infrastructure and the cost-benefit calculus of lead abatement interventions.